Regulatory Rollback and the Smoking Tailpipe: Air-Quality Effects of Repealing Vehicle Emissions Testing

Abstract A growing number of states have repealed mandatory vehicle emissions testing, arguing that advances in vehicle technology and overlapping environmental regulations render such programs redundant while imposing unnecessary costs on drivers. We evaluate this claim by estimating the air-quality effects of emissions-testing repeal using a newly constructed county-by-year dataset of testing requirements linked to U.S. Environmental Protection Agency air-quality data from 1990–2019. Using a staggered policy design, we compare counties that repeal testing to counties that never repeal. We implement three counterfactual-imputation approaches that predict untreated outcomes from pre-repeal trends and contemporaneous untreated counties, allowing us to assess model fit, identifying assumptions, and sensitivity. Our primary matrix-completion specification estimates an average increase in median AQI of 4.7 points, with secondary outcomes showing corresponding shifts in the distribution of air-quality categories. These results are robust across alternative estimators. We evaluate credibility using pre-treatment predictive fit, placebo-style diagnostics, leave-one-out validation, and formal sensitivity analysis that allows limited departures from parallel trends. Overall, emissions-testing repeal is associated with worse air quality on average, with larger effects among earlier repealers and more heterogeneous effects among later cohorts, indicating that nominally "redundant" regulations can remain environmentally consequential. JEL Classification: Q58 , Q53 , C23 , Q52 , K32